• Jul 03, 2017 News!Good News! Since 2017, IJMLC has been indexed by Scopus!
  • Aug 15, 2017 News![CFP] 2017 the annual meeting of IJMLC Editorial Board, ACMLC 2017, will be held in Singapore, December 8-10, 2017.   [Click]
  • Jul 26, 2017 News!The papers published in Vol.7, No.1-3 have all received dois from Crossref.
General Information
Dr. Lin Huang
Metropolitan State University of Denver, USA
It's my honor to take on the position of editor in chief of IJMLC. We encourage authors to submit papers concerning any branch of machine learning and computing.
IJMLC 2016 Vol.6(4): 231-234 ISSN: 2010-3700
DOI: 10.18178/ijmlc.2016.6.4.603

Semi-nonnegative Matrix Factorization Algorithm Based on Genetic Algorithm Initialization

M. Chouh and K. Boukhetala
Abstract—Semi-nonnegative matrix factorization (Semi-NMF) is one of variations of nonnegative matrix factorization model (NMF) when the data matrix X is unconstrained (it may have mixed signs). Semi-NMF decomposes X into two matrices A and B of dimensions n×k and k × p respectively, where each element of the matrix B is nonnegative, such that: X ≈ AB . In the present paper, we proposed a semi-nonnegative matrix factorization algorithm based on genetic algorithm (GA) initialization which has larger searching area and gives the best initialization for the Semi-NMF algorithm to get the optimal solution of semi-nonnegative matrix factorization problem. Also, we compared this initialization for Semi-NMF algorithm with both the random and the k-means initializations introduced in the literature.

Index Terms—Semi-nonnegative matrix factorization, genetic algorithm, initialization.

The authors are with Faculté de mathématiques, USTHB, El-Alia BP 32, Bab-Ezzouar 16111, Alger, Algérie (merich_88@hotmail.com, kboukhetala@usthb.dz).


Cite: M. Chouh and K. Boukhetala, "Semi-nonnegative Matrix Factorization Algorithm Based on Genetic Algorithm Initialization," International Journal of Machine Learning and Computing vol. 6, no. 4, pp. 231-234, 2016.

Copyright © 2008-2015. International Journal of Machine Learning and Computing. All rights reserved.
E-mail: ijmlc@ejournal.net